Genomic markers for decision making:What is preventing us from using markers?

The advent of novel genomic technologies that enable the evaluation of genomic alterations on a genome-wide scale has significantly altered the field of genomic marker research in solid tumors. Researchers have moved away from the traditional model of identifying a particular genomic alteration and evaluating the association between this finding and a clinical outcome measure to a new approach involving the identification and measurement of multiple genomic markers simultaneously within clinical studies. This in turn has presented additional challenges in considering the use of genomic markers in oncology, such as clinical study design, reproducibility and interpretation and reporting of results. This Review will explore these challenges, focusing on microarray-based gene-expression profiling, and highlights some common failings in study design that have impacted on the use of putative genomic markers in the clinic. Despite these rapid technological advances there is still a paucity of genomic markers in routine clinical use at present. A rational and focused approach to the evaluation and validation of genomic markers is needed, whereby analytically validated markers are investigated in clinical studies that are adequately powered and have pre-defined patient populations and study endpoints. Furthermore, novel adaptive clinical trial designs, incorporating putative genomic markers into prospective clinical trials, will enable the evaluation of these markers in a rigorous and timely fashion. Such approaches have the potential to facilitate the implementation of such markers into routine clinical practice and consequently enable the rational and tailored use of cancer therapies for individual patients. © 2010 Macmillan Publishers Limited. All rights reserved.

A decision support system for managing irrigation in agriculture

In this paper, an automatic Smart Irrigation Decision Support System, SIDSS, is proposed to manage irrigation in agriculture. Our system estimates the weekly irrigations needs of a plantation, on the basis of both soil measurements and climatic variables gathered by several autonomous nodes deployed in field. This enables a closed loop control scheme to adapt the decision support system to local perturbations and estimation errors. Two machine learning techniques, PLSR and ANFIS, are proposed as reasoning engine of our SIDSS. Our approach is validated on three commercial plantations of citrus trees located in the South-East of Spain. Performance is tested against decisions taken by a human expert.

An analysis of International NGOs design decision making for housing reconstruction in developing country contexts

The purpose of this paper is to explore the current design decision making process of selected foreign international non governmental organisations (INGO’s) operating in the field of housing and post disaster housing design and delivery in developing countries. The study forms part of a wider on-going study relation to a decision making in relation to affordable and sustainable housing in developing
countries. The paper highlights the main challenges and opportunities in relation to the design and delivery of low cost sustainable housing in developing countries as identified in current literature on the subject. Interviews and case studies with INGO’s highlight any specific challenges faced by foreign INGO’s operating in a developing country. The preliminary results of this research study provide a concise insight into the design decision making process of leading foreign INGO’s operating in developing countries and will be beneficial to policy makers, NGOs, government bodies and community organisations in practice as it offers unique evidence based insights into international bodies housing design decision making process.

Multilinear and Integer Programming for Markov Decision Processes with Imprecise Probabilities

Markov Decision Processes (MDPs) are extensively used to encode sequences of decisions with probabilistic effects. Markov Decision Processes with Imprecise Probabilities (MDPIPs) encode sequences of decisions whose effects are modeled using sets of probability distributions. In this paper we examine the computation of Γ-maximin policies for MDPIPs using multilinear and integer programming. We discuss the application of our algorithms to “factored” models and to a recent proposal, Markov Decision Processes with Set-valued Transitions (MDPSTs), that unifies the fields of probabilistic and “nondeterministic” planning in artificial intelligence research. 

Partially Ordered Preferences in Decision Trees: Computing Strategies with Imprecision in Probabilities

Partially ordered preferences generally lead to choices that do not abide by standard expected utility guidelines; often such preferences are revealed by imprecision in probability values. We investigate five criteria for strategy selection in decision trees with imprecision in probabilities: “extensive” Γ-maximin and Γ-maximax, interval dominance, maximality and E-admissibility. We present algorithms that generate strategies for all these criteria; our main contribution is an algorithm for Eadmissibility that runs over admissible strategies rather than over sets of probability distributions.